From: Larry Bassel <lbassel@codeaurora.org>
To: Andi Kleen <andi@firstfloor.org>
Cc: Larry Bassel <lbassel@codeaurora.org>,
linux-mm@kvack.org, vgandhi@codeaurora.org
Subject: Re: RFC -- new zone type
Date: Fri, 30 Sep 2011 10:01:58 -0700 [thread overview]
Message-ID: <20110930170158.GC7007@labbmf-linux.qualcomm.com> (raw)
In-Reply-To: <m2aa9nhzjf.fsf@firstfloor.org>
On 29 Sep 11 13:19, Andi Kleen wrote:
> Larry Bassel <lbassel@codeaurora.org> writes:
> >
> > It was suggested to me that a new zone type which would be similar
> > to the "movable zone" but is only allowed to contain pages
> > that can be discarded (such as text) could solve this problem,
>
> This may not actually be a win because if the text pages are needed
> afterwards the act of rereading them from disk would likely take longer
> than the copying.
Yes, I'm aware of this (they'd actually be coming from flash
which is even slower, right?).
>
> The so you many not get latency before, but after.
>
> Essentially robbing Peter to pay Paul.
Yes, the goal is to create this large contiguous memory
quickly, even if performance is worse later on for a while.
>
> If the goal is to just spread the latency over a longer time
> I'm sure there are better ways to do that than to add a new zone.
I myself am not an advocate of creating this new zone (because
I think there may not be a lot of benefit to it for the
amount of work involved), but I want to solicit people's
opinions about it and collect suggestions (such as Dan's)
about alternate approaches that may be better/closer
to already implemented (I'd likely have to port them
to ARM as they probably were developed for x86, but I don't
think that's a problem) that will handle our use case.
I wonder if another reasonable approach might be a combination
of memory compaction in the background when the machine
is otherwise idle, combined with occasional killing of
processes that haven't been used for a long time (this is
going to be run on an android user environment, android
likes to leave lots of unused processes running in the
background).
>
> -Andi
Larry
--
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
prev parent reply other threads:[~2011-09-30 17:02 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-09-28 18:09 RFC -- new zone type Larry Bassel
2011-09-29 6:07 ` Sameer Pramod Niphadkar
2011-09-29 16:38 ` [Xen-devel] " Dan Magenheimer
2011-09-29 17:00 ` Larry Bassel
2011-10-05 16:56 ` Larry Bassel
2011-10-05 19:43 ` Dan Magenheimer
2011-10-06 23:03 ` Larry Bassel
2011-10-07 15:23 ` Dan Magenheimer
2011-10-07 16:01 ` Seth Jennings
2011-10-07 17:19 ` Larry Bassel
2011-10-12 22:20 ` Dan Magenheimer
2011-09-29 20:19 ` Andi Kleen
2011-09-30 17:01 ` Larry Bassel [this message]
[not found] <20110928180909.GA7007@labbmf-linux.qualcomm.comCAOFJiu1_HaboUMqtjowA2xKNmGviDE55GUV4OD1vN2hXUf4-kQ@mail.gmail.com>
[not found] ` <c2d9add1-0095-4319-8936-db1b156559bf@default20111005165643.GE7007@labbmf-linux.qualcomm.com>
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20110930170158.GC7007@labbmf-linux.qualcomm.com \
--to=lbassel@codeaurora.org \
--cc=andi@firstfloor.org \
--cc=linux-mm@kvack.org \
--cc=vgandhi@codeaurora.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).